Carlsbro 50Top bias problem

[todor]

Hello everybody!

I have a 50Top sitting on my bench, new tubes, cap job done. And while I was at it, I also installed an adjustable bias, but noticed that the negative voltage doesn't get high (or low) enough. I only get about -41V at most and the tubes (EL34s) are running at about 40mA @ 480V each, which is a bit too much going by the 70% rule. I'd like to get a safe range for adjustement.

The separate bias winding should give -60V after rectified, which it does when I disconnect the shunt resistance, leaving only the series resistor. After the diode there's only a 4.7k series resistor and (originally) a 47k shunt resistor parallelled by 47uF. And for some reason adding any load to the circuit drags the voltage down to -40V. Before diode I measure about 42Vac at all times.

All this happens on standby. And I already replaced the diode in the circuit. Any other ideas?

Thanks in advance!

-Tommi

[Jazz P Bass]

Where is the "adjustable bias" part of the circuit?
You stated a 4.7K series resistor & a 47K shunt R.
I don't see "adjustable.
Did you try a higher shunt R value?

[todor]

I originally replaced the shunt resistor with a 22k trimmer and a 33k resistor, but changed the values back to original to ease the troubleshooting. And I also increased the shunt resistance value but it didn't affect the situation in any way, neither did changing the series resistance.

[MWJB]

Replacing the 47K load resistor with a pot totalling 55K isn't going to get you anywhere (0.91 ratio, compared to 0.92). You know you need at least 55K at the load, so fit a 56K resistor to ground, plus a 50K trim pot in series, wired as a variable resistor. You can always reduce the value of the 4.7K dropper to bring up the neg voltage a little (trim pot sweep will be reduced).

I'm also wondering whether there isn't an electrolytic cap missing from between the bias diode & the 4.7K?

Is the amp solid state rectified? If a tube, do you have scope to try a lower voltage rectifier, or a higher wall AC tap (watch heater vltages don't drop below 6.3VAC)?

[todor]

Replacing the 47K load resistor with a pot totalling 55K isn't going to get you anywhere (0.91 ratio, compared to 0.92). You know you need at least 55K at the load, so fit a 56K resistor to ground, plus a 50K trim pot in series, wired as a variable resistor. You can always reduce the value of the 4.7K dropper to bring up the neg voltage a little (trim pot sweep will be reduced).

I'm also wondering whether there isn't an electrolytic cap missing from between the bias diode & the 4.7K?

Is the amp solid state rectified? If a tube, do you have scope to try a lower voltage rectifier, or a higher wall AC tap (watch heater vltages don't drop below 6.3VAC)?

But even with the original values the voltage seems a bit too positive? And the voltage drops right after the diode. I'll try reducing the series dropping resistor.

The only filter cap is the 47uF, in the schematic also, and there's no spot on the PCB for a missing cap. Schemes: lCarlsbro 50 Top Amp Schematic

[Jazz P Bass]

Let me see if I have this correct.
Without any shunt to ground, you are reading -60Vdc where?
At the grids?
And with a 47K to ground it "drops" to -40Vdc.
The grids alone should not be pulling current enough to "drop" the bias voltage.
I did a voltage divider calculation : 4.7K & 47K should read -54.5Vdc.
Link:Voltage Divider Calculator

[MWJB]

Looks like they ran them quite hot back in the day...doesn't translate welll to today's higher wall AC. So, yes, original values are probably not negative enough. Even if you drop the 4.7K to 1K, or less you still won't get an appropriate rise in negative voltage, you need to increase the load resistor too, or convert to a voltage doubler bias circuit...or fit colder running 6550 tubes (Sovtek?) that will cope with big currents.

[todor]

Let me see if I have this correct.
Without any shunt to ground, you are reading -60Vdc where?
At the grids?
And with a 47K to ground it "drops" to -40Vdc.
The grids alone should not be pulling current enough to "drop" the bias voltage.
I did a voltage divider calculation : 4.7K & 47K should read -54.5Vdc.
Link:Voltage Divider Calculator

Without the shunt res I read about -57V after the series resistor, on standby, that is. And after adding the shunt (virtually any value...) the whole rail drops to about -40Vdc.

I'll try to lower the series resistance and change the ratio as far as it goes like MWJB suggested... I'll get back on it tomorrow.

[Jazz P Bass]

Pull the output tubes.
Measure the Vdc right at the grid pins.
The schematic indicates -30Vdc.
(sounds kind of hot)

[todor]

Hi!

I lowered the series resistance and fiddled with the shunt values and got the negative voltage right. I guess it was just supposed to be that hot originally. I was too focused on the "problem" to even have notived that there was a marked grid voltage for output tubes in the schematic. :P

But problem solved, thank you guys for the input!!

-Tommi

[Jazz P Bass]

Just curiuos what value resistors you ended up with & what bias voltage.
That -30V certainly got my attention.